Appendix C
Quadratic Reciprocity: an
alternative proof
Hundreds of different proofs of this theorem have been published. Gauss,
who first proved the result in 1801, gave eight different proofs. We gave a
group-theoretic proof in chapter 10. Here is a shorter combinatorial proof.
Theorem C.1. (The Law of Quadratic Reciprocity) Suppose p, q ∈ N are
odd primes. Then
(
p−1 q−1
−1 if p ≡ q ≡ 3 mod 4
p q
= (−1) 2 2 =
q
p
1
otherwise.
Proof. Let
q−1
p−1
}, T = {1, 2, . . . ,
}.
2
2
We shall choose remainders modp from the set
S = {1, 2, . . . ,
p
p
{− < i < } = −S ∪ {0} ∪ S,
2
2
and remainders modq from the set
q
q
{− < i < } = −T ∪ {0} ∪ T.
2
2
By Gauss’ Lemma,
q
p
µ
= (−1) ,
= (−1)ν .
p
q
Writing #X for the number of elements in the set X,
µ = #{i ∈ S : qi mod p ∈ −S}, ν = #{i ∈ T : pi mod q ∈ −T }.
By ‘qi mod p ∈ −S’ we mean that there exists a j (necessarily unique)
such that
qi − pj ∈ −S.
But now we observe that, in this last formula,
0<i<
p
q
=⇒ 0 < j < .
2
2
3–1
R3
R1
R2
R4
Figure C.1: p = 11, q = 7
The basic idea of the proof is to associate to each such contribution to µ
the ‘point’ (i, j) ∈ S × T . Thus
p
µ = #{(i, j) ∈ S × T : − < qi − pj < 0};
2
and similarly
q
ν = #{(i, j) ∈ S × T : 0 < qi − pj < },
2
where we have reversed the order of the inequality on the right so that both
formulae are expressed in terms of (qi − pj).
Let us write [R] for the number of integer points in the region R ⊂ R2 .
Then
µ = [R1 ], ν = [R2 ],
where
q
p
R1 = {(x, y) ∈ R : − < qx − py < 0}, R2 = {(x, y) ∈ R : 0 < qx − py < },
2
2
and R denotes the rectangle
p
p
R = {(x, y) : 0 < x < , 0 < y < }.
2
2
The line
qx − py = 0
is a diagonal of the rectangle R, and R1 , R2 are strips above and below the
diagonal (Fig C).
This leaves two triangular regions in R,
p
q
R3 = {(x, y) ∈ R : qx − py < − }, R4 = {(x, y) ∈ R : qx − py > }.
2
2
3–2
We shall show that, surprisingly perhaps, reflection in a central point
sends the integer points in these two regions into each other, so that
[R3 ] = [R4 ].
Since
R = R1 ∪ R2 ∪ R3 ∪ R4 ,
it will follow that
[R1 ] + [R2 ] + [R3 ] + [R4 ] = [R] =
p−1q−1
,
2
2
ie
µ + ν + [R3 ] + [R4 ] =
p−1q−1
.
2
2
But if now [R3 ] = [R4 ] then it will follow that
µ+ν ≡
p−1q−1
mod 2,
2
2
which is exactly what we have to prove.
It remains to define our central reflection. Note that reflection in the
centre ( p4 , 4q ) of the rectangle R will not serve, since this does not send integer points into integer points. For that, we must reflect in a point whose
coordinates are integers or half-integers.
We choose this point by “shrinking” the rectangle R to a rectangle bounded
by integer points, ie the rectangle
R0 = {1 ≤ x ≤
q−1
p−1
, 1≤y≤
}.
2
2
Now we take P to be the centre of this rectangle, ie
P =(
p+1 q+1
,
).
2
2
The reflection is then given by
(x, y) 7→ (X, Y ) = (p + 1 − x, q + 1 − y).
It is clear that reflection in P will send the integer points of R into
themselves. But it is not clear that it will send the integer points in R3 into
those in R4 , and vice versa. To see that, let us shrink these triangles as we
shrank the rectangle. If x, y ∈ Z then
qx − py < −
p
p+1
=⇒ qx − py ≤ −
;
2
2
and similarly
qx − py >
q
q+1
=⇒ qx − py ≥
.
2
2
3–3
Now reflection in P does send the two lines
p+1
q+1
qx − py = −
, qx − py =
2
2
into each other; for
qX − pY = q(p + 1 − x) − p(q + 1 − y) = (q − p) − (qx − py),
and so
p+1
p+1
q+1
⇐⇒ qX − pY = (q − p) +
=
.
2
2
2
We conclude that
[R3 ] = [R4 ].
qx − py = −
Hence
[R] = [R1 ] + [R2 ] + [R3 ] + [R4 ] ≡ µ + ν mod 2,
and so
µ + ν ≡ [R] =
p−1q−1
.
2
2
Example: Take p = 37, q = 47. Then
37
47
=
since 37 ≡ 1 mod 4
47
37
10
=
37
2
5
=
37 37
5
since 37 ≡ −3 mod 8
=−
37
37
=−
since 5 ≡ 1 mod 4
5
2
=−
5
= −(−1) = 1.
Thus 37 is a quadratic residue mod47.
2
We could have avoided using the result for
:
p
10
−27
=
37
37
3
−1
3
=
37
37
37
= (−1)18
3
1
=
= 1.
3
3–4